Conditions for Confidence Interval for p
The method of sampling required for a confidence interval for p.
What is a random sample or simple random sample (SRS) from the population of interest?
The margin of error only accounts for this.
What is sampling variability?
Critical value for a 95% 1 sample z interval for a proportion rounded to two decimal places.
What is 1.96?
If n = 100 and p-hat = .3, what is the standard error to four decimal places.
What is .0458?
The slope of a linear regression, b, can be calculated with this formula.
What is b = r*sy/sx ?
To develop a confidence interval for the difference in two proportions p1 - p2 , 3 condition must be met. They are:
What is (1) random - data come from two independent random samples or two groups in a randomized experiment, (2) 10% condition when sampling without replacement n1 < .10 N1 and n2 < .10 N2 (if an experiment, randomly assign treatments to both groups, (3) Large Counts - at least 10 successes and 10 failures in both groups.
nhatp and n(1 - hatp)
What is the large counts condition?
The MOE for a confidence interval for a proportion is this (give the formula).
What is hat p +-sqrt((hatp(1-hatp))/n
Critical value for a 90% 1 sample z interval for a proportion rounded to three decimal places.
What is 1.645?
The standard deviation is a) larger or b) smaller if you were to sample without replacement and if you were to sample more than 10% (which we're not going to do).
What is smaller?
To linearize data, if (log x, log y) linearizes the data then a ___________ model is appropriate.
What is a Power Model?
For #8.62 on p. 574, answer the question.
What is (a) random - yes, crackers were assigned to one of two treatment groups, (b) The 10% condition does not apply because no sampling took place - this was an experiment with random assignment of treatments, (c) large counts - not met - there are fewer than 10 successes (0 in the the microwave group)?
To assure that the sampling distribution is approximately Normal we need the...
What is the large counts (Normality Condition). n*p greater or = to 10 and n*(1-p) greater or = 10.
True or False: MOE includes response and nonresponse biases.
What is FALSE?
Critical value for a 99% 1 sample z interval for a proportion rounded to three decimal places.
What is 2.576?
When sampling without replacement, we check the 10% condition (sample less than 10% of the population), so that __________.
What is to account for the fact that we are sampling without replacement and the observations are actually not independent, so by sampling less than 10% of the population are standard deviation (error) is approx. correct, so our standard deviation (standard error) can be used fairly accurately in Normal calculations?
The least squares regression line always goes through ______.
What is
(barx, bary)?
For #64 p. 574, answer the question:
What is (a) random - independent random samples were taken - condition met; (b) 10% condition - yes, 500 is less than 10% of all DVDs produced by each machine, (c) Large Counts - 480, 20, 484, 16 are all greater than or = to 10 - condition met?
When sampling without replacement, this condition is needed.
What is the 10% condition. n < .10N meaning, you must sample less than 10% of the population.
To cut the MOE in half, do this to n (the sample size).
What is multiply the sample size by 4?
Critical value for a 86% 1 sample z interval for a proportion rounded to two decimal places.
What is 1.48?
If n = 100 and p-hat = .4, what is the standard error to three decimal places.
What is .049?
P (A/B) Give definition
(P (AnnB))/(P(B))?
For #72 p. 575, answer letter (a).
What is we do NOT have convincing evidence that the two population proportions are different because the interval includes 0 as a plausible value for the difference (rural - urban) in the true proportion of households that have a natural tree?
If your sample is a voluntary response sample or a convenience sample and you are going to construct a confidence interval for p, you must make this/these adjustment(s).
What is no adjustments can be made. You must trash your sample results and start over with an SRS from the population of interest.
2 ways to shorten (or reduce) a confidence interval.
What are 1) increase the sample size and 2) lower the confidence level?
Critical value for a 94% 1 sample z interval for a proportion rounded to two decimal places.
What is 1.88?
If were were determining the sample size, n, needed to be sure that the margin of error for your 1 sample 95% confidence interval for a population proportion was less than 2 % (.02) and no pilot proportion was given, you would use (a) what value for p-hat and b) what is the sample size needed.
a) What is use .5 for p-hat and b) What is n = 2,401?
Assuming this qualifies for a binomial distribution (does it?), what is the probability Al scores more than 2 goals in 10 shots to four decimal places?
What is .3222?
For #72 p. 575 answer letter (b):
What is no? Because 0 is captured in the interval, it is plausible that the population proportions are equal, but this does NOT provide convincing evidence that the two proportions are equal.